I believe this and based on anecdotal evidence this seems to be an inherent difference between males and females. I’ll theorize that it’s because, from an evolutionary biology perspective, women are much more dependent on the community to keep themselves and their children safe so it’s more important for them to closely align with others in the community.
That said, I don’t know how to apologize on behalf of my sex. This is so wrong. The answer to speech we disagree with isn’t banning the speech, it’s more speech making the argument for why the initial speech was wrong. Debate is how we arrive at the best ideas; we don’t just accept the first thing someone says.
In my eyes, there's no need to apologize on behalf of anyone else. You're you! You aren't and cannot be responsible for other women's actions, only your own. Thus, you have nothing to apologize for.
agreed, she shouldn't applogize! it's interesting i have never seen women respond this graciously to men professing guilt on behalf of their gender. only some variation of "yeahhh, not looking great. you gotta do better and police other men." differences indeed.
These remarks are always revealing. I have explained to women that they are essentially asking for fascism. They get very angry, but the more intelligent ones learn.
You’re all right that it doesn’t make sense for a person to apologize for other people they share nothing with other than demographics. I just found myself so disappointed in other women!
Although there may be some inherent differences and women may be wired to value group alignment more than men, it’s still definitely possible for women who are educated in the right environment to value free speech and debate. I am a very conflict-adverse person by nature, but I still truly don’t understand how anyone can think it’s a good idea to disallow speech you disagree with. This goes against every principle of a free society!
When some man says “I apologise on behalf on my gender (or race, or …)” my response is not “how noble”. It’s “I didn’t ask you to speak in my behalf, you snivelling git”. This is true even if I think they have a point. Apologise on your own behalf. Act to make changes. But don’t appoint yourself spokesman for others - even if you agree that your demographic tends towards certain behaviours
This line of thought leads to the evil of collective guilt. By analogy: Collective caution is fine; young people as a whole pay higher car insurance premiums because as a whole they are riskier drivers. But if one young person causes an accident then they alone should be held accountable - it’s neither just or constructive to demand an accounting or contrition from those who were uninvolved.
I may have chosen my words poorly. You may also be reading way too much into it.
I strongly agree with you that we are all individuals, not representatives of demographic groups, and that thinking all members of demographic groups are interchangeable is part of what’s gotten us in the mess we’re in.
I certainly don’t believe I speak for all women on the planet, and it was admittedly an attempt to distance myself from a behavior I find abhorrent but that some people would assume I am somehow inherently inclined to engage in based on the statistics being presented. I think it’s natural to feel a little dismayed when confronted with credible evidence that a demographic you’re part of engages in a harmful behavior at higher rates. Especially when it’s a demographic like sex that DOES drive some inherent behavioral differences.
All that said, I think your reaction was a bit much, but you’re entitled to believe I’m a sniveling git if you like.
My first paragraph was specifically about men, and how many men feel when someone else presumes to represent them.
Does it also apply to women? I don’t know. I hear your words as coming from a place of sympathy - “these men have been hurt by poor behaviour of some women & I feel for them.” But most men don’t want you to lower yourself by trying to carry someone else’s guilt
You shouldn't be responsible for apologizing for people aside from yourself. I'm not sure apologies are required anyway for an inherent fact of biology. Women have been in the forefront of religion for a very long time, even back as far as early Christian practice, and maybe before that, we don't have a lot of evidence. How women would be treated in the Church was the subject of a lot of strife in the pre-Constantine church, since women were so prominent. Dogma and belief are at cross purposes to free expression. Your theorizing is probably pretty close to the mark.
I suspect you're right, but you can't tell from a single survey in a single time and place nurture versus nature. You might start to get at that if you repeated the survey and saw the same pattern in China, Brazil, and the Congo.
And I wouldn't get too guilty. You can't blame a random guy on the street for things other men have done...but you might tell your daughter if a guy invites her over 'to play video games', he might have something else than that in mind.
And if you stay with the evolutionary perspective, maybe men were more often placed in positions where negotiation with another viewpoint might prevent violence?
The prime example of the difference is that you'd feel the need to offer some sort of apology for being female. Acting as such is definitely not a male trait.
You are probably right that more women than men would feel this way. But certainly there are many examples of men attempting to distance themselves from the behavior of other men and paint themselves as “one of the good guys”. It is human nature to try to avoid criticism.
Tolerance of other views (hearing them without screaming or trying to get the speaker fired, etc.) correlates strongly with the level of confidence the listener has in his own views, I think. Anyone who is secure in his own beliefs and actions has no need to fear contrary perspectives, and is content to answer them with reason, not force.
I think you're right to an extent, but I also think it has to do with curiosity and open-mindedness. As a woman and someone who is not entirely confident in my views about any one issue (I am confident in my principles, though), I would never shout down or try to cancel a speaker because I know that by allowing them, as well as their opponents, to speak, my views will be informed and perhaps more solidified as a result.
Another essential quality is intellectual humility, recognizing that you cannot expect to understand everything, or even just another human being, so you know that there is value in listening to people with a different perspective.
Hi Chapin! Fascinating stuff but could you point me to where the raw data are? I couldn't find them at the link you gave, and the heat maps are frankly giving me a headache :)
The tolerance values are calculated as even sums of the left/right speaker questions and scaled 0-100%. Let me know if you have other questions about the data/my analysis.
Tolerance is measured by asking whether students would allow hypothetical controversial speakers to speak on campus. It's explained in the longer article, and the statements are listed in the addendum:
Most of the American cancel culture participants are nothing more than the modern equivalent of spinster church ladies. They love to point out the flaws of others and point out their own virtue. In the absence of public morality they now use more fashionable metrics such as bigotry. This is most obvious in the "me-too" accusations where they struggle so hard because consent is the only concept left to exploit.
Now that people stopped attending church, the busy bodies moved on to becoming university administrators and high school teachers. They LOVE judging others. They ruin every industry they come to dominate, then blame everyone else for running away. Even the women who hate this have trouble recognizing it for what it is. instead they say "I hate working for women" when in reality they hate being around close-minded hypocrites who happen to be female.
The only way to fix this is to legislate it, and to promote rules-based environments while dismissing and demonizing taboo-based environments. The reason so many places are so oppressive is because these busy-bodies love to enforce taboos, and they hate rules because they are explicit and cannot be warped in order to attacks others.
Tolerance is essential in a multicultural society. We are careening towards civil war. We must fix broken environments before the rot spreads to important institutions.
Thankfully we have organizations like Braver Angels and BridgeUSA that are working to cultivate tolerance and respectful disagreement. I just wrote Substack articles about both of their efforts; I hope they become more mainstream.
The fact that only one of the issues directly affected women doesn't exactly jam up that argument. It would also make sense if, being historically more constrained by the consequences of political decisions, women are generally more reactive on the behalf of whatever group is most affected while men get to remain aloof and treat political discourse as a kind of game. I suspect this is at least part of a larger explanation.
I think women tend to value feelings/emotions more. They don't wanna see people get their feelings hurt. And some opinions hurt people's feelings, sooo...
We have rules in place to deal with male rage and violence. We do not have any structures or policies in place to deal with female shaming and reputation destruction. Besides, male violence peaks in their teens and is no longer needed once they are out in a world where one's status is defined.
Men are no more violent today than they were in the past. Our rules can handles that, at least when we want them to. Of course, now many cities have decided to stop punishing violent men. Want to guess which sex was pushing for that change?
"Men are no more violent today than they were in the past"
And if anything men in Europe and east Asia are far less violent than in the past because we selectively executed about 1% of the male population each generation (largely psychopaths) for a very very long time (tribal societies did not do this, they instead rewarded the most violent men with harems).
I do not understand your comment. Are you saying we should stop enforcing crime? Should we legalize rape and domestic violence? I cannot understand where your comment came from, unless it is a bot.
If the difference is simply that women prefer more social harmony, wouldn’t we expect to see this effect across generations? Has anyone done that study? I’d bet Gen X and Boomers have less of this effect and if we could find enough of the Greatest Generation still, we wouldn’t see this gap at all, or it might not even reach the level of significance. There are other things going on with millennial and Gen Z women: they have been educated by teachers and professors to despise debate and have been overwhelmingly captured by those views, while those views have actually driven men away, in large part because the package of views includes the concept of “toxic masculinity”. Another potential contributor to the effect could be that young women are far more engaged on social media than young men are. Young men are more engaged with online video games. Social media literally trains your brain to react in strongly antisocial ways to things you disagree with, due to the internal reward system of clicks, likes, and shares.
We see the gap in all generations. Female professors and students are both far more likely to support censorship in academia than their male equivalents.
Women were also the staunchest supporters of totalitarian movements in other countries.
But both the professors and the students only make it into their jobs and placements if they parrot a suite of leftist views which includes support for censorship. So a sample of only people involved in academia is extremely skewed. In order to determine whether women are truly less tolerant in significant percentages, a good study would include a broader sample of men and women from across society, not just from academia.
Women did not dominate certain fields in the past. We are in a brave new world now where industries such as book publishing, journalism, university administration and others are now dominated by women for the first time in human history. It is in these fields where we see the decline in tolerance.
Interesting evidence to a fairly common anecdotal claim.
I've theorized for some time now that this is potentially linked to direct experience of the consequences of physical violence in interpersonal disputes.
Little boys grow up with much more experience in this area than little girls. Hence, they think differently about the consequences of escalating conflict. Women, being more inclined to covert or indirect forms of violence, rarely feel the sting of retributive conflict escalation. Engaged by proxy, it will be men who bare the burden of this socially.
Your plotted outcomes appear to belie the idea that there is no difference between left and right on this - after controlling for sex, the right-wing show more tolerance than the left (according to the plot, but not sure if this is statistically significant). This effect is smaller than the sex effect, but it looks like there is some effect. You should investigate and revise your article accordingly.
I took some time to look into the analysis underlying your post because I wanted to understand precisely how the claim that “men are more than three and a half times more likely to be perfectly tolerant than women” was derived.
To do so, I read the subsequent discussion in the comments in which you indicated the data source and the methodology, and downloaded and re-analyzed the publicly available raw data from the 2026 College Free Speech Rankings that you used.
One suggestion up front: I think the methodology should be set out explicitly in the article itself rather than being left to the comments section. I spent a considerable amount of time trying to reproduce the result, and without the methodological clarification it is very difficult to understand how the headline claim is obtained from the underlying data.
My understanding is that the analysis is based on responses of men and women to the following survey item:
"Student groups often invite speakers to campus to express their views on a range of topics. Regardless of your own views on the topic, should your school ALLOW or NOT ALLOW a speaker on campus who has previously expressed the following idea?"
The six statements are:
"Transgender people have a mental disorder."
"Abortion should be completely illegal."
"Black Lives Matter is a hate group."
"The Catholic Church is a pedophilic institution."
"The police are just as racist as the Ku Klux Klan."
"Children should be able to transition without parental consent."
I downloaded the raw male and female datasets, applied the survey weights (School Universe Weight), excluded missing values, and calculated support as: "Probably should allow" + "Definitely should allow"
Using that methodology, I obtained the following results:
Statement Women allowing (%) Men allowing (%) Difference (percentage points)
"Transgender people have a mental disorder." 17.5 39.1 +21.6
"Abortion should be completely illegal." 30.9 53.9 +23.0
"Black Lives Matter is a hate group." 15.6 36.7 +21.1
"The Catholic Church is a pedophilic institution." 31.0 42.3 +11.3
"The police are just as racist as the Ku Klux Klan." 30.3 40.8 +10.5
"Children should be able to transition without parental consent." 46.3 48.7 +2.4
These results suggest that men are consistently more likely than women to allow the speakers in question, but that the size of the gap varies substantially by topic, from 2.4 to percentage points to 23 in the case of someone having a track record of supporting the full ban of abortion.
What puzzled me is that I could not reproduce the "3.5 times" finding using what I understand to be the actual tolerance scale. If tolerance is calculated, as described, by combining the speaker questions into an equally weighted index and scaling it from 0 to 100, I obtain approximately:
Women: 33.6
Men: 45.9
In other words, men score about 12 points higher than women on the tolerance scale, or roughly 1.37 times higher, not 3.5 times higher.
The only way I was able to reproduce a figure close to the one reported in the article was by constructing a separate category of "perfect tolerance", defined as respondents who answered "Definitely should allow" on all six questions simultaneously. Under that definition, I obtain approximately:
Women: 1.9%
Men: 6.8%
which yields a ratio of about 3.6:1.
If that is indeed the methodology used, I think it is essential that it be stated prominently in the article. The phrase "perfectly tolerant" strongly implies that the result is a direct measure of tolerance, whereas in reality it depends entirely on a very specific coding decision requiring the maximum possible response on six separate questions.
My concern is that the article currently invites readers to conclude that the data show men to be generally 3.5 times more tolerant than women. However, the underlying data seem to support a much narrower and more nuanced conclusion: men are more likely than women to allow the specific speakers included in the survey, with the magnitude of the difference varying considerably across issues.
Put differently, the headline appears to rest not on the tolerance scale itself, but on a highly restrictive binary classification applied afterward. Because only 6.8% of men and 1.9% of women satisfy that criterion, the resulting ratio is largely a product of that methodological choice rather than a straightforward summary of the underlying data.
Could you thus kindly clarify whether the 3.5x figure was derived from the requirement that respondents answer "Definitely should allow" on all six items, or how else you came to this finding.
Given how central these choices are to the article's conclusion, I think greater methodological transparency would substantially strengthen the analysis and help readers evaluate what is actually being measured.
Thanks for your comment. Yeah, that's exactly how I got the 3.5 figure.
In my articles, I'm always balancing trying to be as detailed and transparent about my methodology as possible and trying to be accessible to people without technical backgrounds. Not as easy task, and I'm always open to suggestions on how to do so better. I would've guessed that saying "men are more than three and a half times more likely to be perfectly tolerant than women" would be fairly clear. (I can't think of anything way to interpret "perfectly tolerant" besides "would definitely allow all speakers". Can you?) Though I've seen others misinterpret that line as meaning "3.5x more tolerant".
Also, the 3.5 figure, while notable, isn't central to the article's main conclusions. The central result of the article is that men often have grater tolerant towards speakers on the other side than women do towards speakers on their own side, which comes from the findings that slightly conservative men being more tolerant of left-wing speakers than democratic socialist women, a slightly liberal men being more tolerant of right-wing speakers than somewhat conservative women, etc.
See the longer version of the article, which notably has heatmaps where perfect tolerance is the top right square:
I am going to say it. Nowhere is this more glaring than with the issue of abortion. When the word is spoken, the sword drops on the table dividing women, erasing men and galvanizing rhetoric. Unfortunately for many women, their feet are in concrete on this issue. It becomes the base of their political views that they can not be moved from it.
Exactly this is what makes them poor politicians, and citizens that base their vote on one issue.
This essay presents itself as empirical clarity, but it repeatedly turns a narrow survey result into a moral judgment the data does not support.
What the data actually shows is straightforward: male and female students answer a specific survey question differently. Men are more likely to say they would allow controversial speakers on campus; women are more likely to say they would not. This pattern holds across political ideology, and the gender gap is larger than the ideological gap. That is the finding.
Everything beyond that is interpretation.
The survey measures tolerance as willingness to allow hypothetical speakers in an abstract, consequence-free setting. It does not measure openness to debate, reasoning ability, emotional stability, or authoritarian impulse. It does not measure how people behave when speech has real institutional, social, or personal costs attached. Treating this metric as a proxy for character is unjustified.
The essay nonetheless frames male responses as civic virtue and female responses as censorial instinct. That move is emotional, not empirical.
From the beginning, tolerance is treated as an unqualified good and refusal as a failure. But the data never weighs error costs. It never asks whether some refusals are rational, preventative, or responsive to real-world asymmetries. Those assumptions are simply smuggled in.
The dismissal of alternative explanations is similarly weak. The author argues that women aren’t responding differently because they’re more affected by speech, since only abortion is explicitly sex-linked and the gap doesn’t spike there. This misunderstands how risk works. Speech-related costs for women aren’t limited to specific policy topics; they include higher exposure to harassment, reputational penalties, social sanction, and uneven institutional protection. These are ambient conditions, not speaker-specific variables, and the survey cannot detect them.
There’s also a clear narrative asymmetry. Men’s responses are treated as the neutral baseline. Women’s responses are treated as a deviation that requires explanation. That is not a methodological necessity; it’s a framing choice.
The most revealing moment is the speculative turn: the suggestion that women “prefer social harmony.” Presented as modest conjecture, it functions as character inference. Harmony becomes conflict aversion, which becomes censorship, which becomes a threat to democratic discourse. None of those steps are supported by the data. They are layered assumptions.
What’s entirely missing is any discussion of the costs of excessive tolerance. The essay treats tolerance as costless and censorship as uniquely dangerous. But tolerating speech and governing institutions are not the same task. Universities are not abstract marketplaces of ideas; they are organizations with duty-of-care obligations and power asymmetries. A respondent who weighs those realities more heavily will score as “less tolerant” by design.
By the end, the essay stops analyzing data and starts issuing moral warnings about emotional comfort undermining society. At that point, the data is no longer being interpreted; it’s being recruited to serve a worldview.
A more accurate summary would be simple: men and women answer abstract questions about allowing controversial speakers differently, and this difference outweighs ideological variation. Explaining that gap requires assumptions about harm, risk, and institutional responsibility that the data itself cannot resolve.
The data is real. The emotional story built on top of it is not.
I believe this and based on anecdotal evidence this seems to be an inherent difference between males and females. I’ll theorize that it’s because, from an evolutionary biology perspective, women are much more dependent on the community to keep themselves and their children safe so it’s more important for them to closely align with others in the community.
That said, I don’t know how to apologize on behalf of my sex. This is so wrong. The answer to speech we disagree with isn’t banning the speech, it’s more speech making the argument for why the initial speech was wrong. Debate is how we arrive at the best ideas; we don’t just accept the first thing someone says.
In my eyes, there's no need to apologize on behalf of anyone else. You're you! You aren't and cannot be responsible for other women's actions, only your own. Thus, you have nothing to apologize for.
agreed, she shouldn't applogize! it's interesting i have never seen women respond this graciously to men professing guilt on behalf of their gender. only some variation of "yeahhh, not looking great. you gotta do better and police other men." differences indeed.
These remarks are always revealing. I have explained to women that they are essentially asking for fascism. They get very angry, but the more intelligent ones learn.
You’re all right that it doesn’t make sense for a person to apologize for other people they share nothing with other than demographics. I just found myself so disappointed in other women!
Although there may be some inherent differences and women may be wired to value group alignment more than men, it’s still definitely possible for women who are educated in the right environment to value free speech and debate. I am a very conflict-adverse person by nature, but I still truly don’t understand how anyone can think it’s a good idea to disallow speech you disagree with. This goes against every principle of a free society!
Appreciate your willingness to discourse on this
When some man says “I apologise on behalf on my gender (or race, or …)” my response is not “how noble”. It’s “I didn’t ask you to speak in my behalf, you snivelling git”. This is true even if I think they have a point. Apologise on your own behalf. Act to make changes. But don’t appoint yourself spokesman for others - even if you agree that your demographic tends towards certain behaviours
This line of thought leads to the evil of collective guilt. By analogy: Collective caution is fine; young people as a whole pay higher car insurance premiums because as a whole they are riskier drivers. But if one young person causes an accident then they alone should be held accountable - it’s neither just or constructive to demand an accounting or contrition from those who were uninvolved.
I may have chosen my words poorly. You may also be reading way too much into it.
I strongly agree with you that we are all individuals, not representatives of demographic groups, and that thinking all members of demographic groups are interchangeable is part of what’s gotten us in the mess we’re in.
I certainly don’t believe I speak for all women on the planet, and it was admittedly an attempt to distance myself from a behavior I find abhorrent but that some people would assume I am somehow inherently inclined to engage in based on the statistics being presented. I think it’s natural to feel a little dismayed when confronted with credible evidence that a demographic you’re part of engages in a harmful behavior at higher rates. Especially when it’s a demographic like sex that DOES drive some inherent behavioral differences.
All that said, I think your reaction was a bit much, but you’re entitled to believe I’m a sniveling git if you like.
Hey Dee. Thanks for responding.
My first paragraph was specifically about men, and how many men feel when someone else presumes to represent them.
Does it also apply to women? I don’t know. I hear your words as coming from a place of sympathy - “these men have been hurt by poor behaviour of some women & I feel for them.” But most men don’t want you to lower yourself by trying to carry someone else’s guilt
You shouldn't be responsible for apologizing for people aside from yourself. I'm not sure apologies are required anyway for an inherent fact of biology. Women have been in the forefront of religion for a very long time, even back as far as early Christian practice, and maybe before that, we don't have a lot of evidence. How women would be treated in the Church was the subject of a lot of strife in the pre-Constantine church, since women were so prominent. Dogma and belief are at cross purposes to free expression. Your theorizing is probably pretty close to the mark.
I mean, we don't know that from this data.
I suspect you're right, but you can't tell from a single survey in a single time and place nurture versus nature. You might start to get at that if you repeated the survey and saw the same pattern in China, Brazil, and the Congo.
And I wouldn't get too guilty. You can't blame a random guy on the street for things other men have done...but you might tell your daughter if a guy invites her over 'to play video games', he might have something else than that in mind.
True. And I don’t feel guilty, just dismayed.
The cross-cultural examples exist. Not for Congo though.
And if you stay with the evolutionary perspective, maybe men were more often placed in positions where negotiation with another viewpoint might prevent violence?
The prime example of the difference is that you'd feel the need to offer some sort of apology for being female. Acting as such is definitely not a male trait.
Biology wins again.
You are probably right that more women than men would feel this way. But certainly there are many examples of men attempting to distance themselves from the behavior of other men and paint themselves as “one of the good guys”. It is human nature to try to avoid criticism.
Most of the time when men do it, is it men pretending to be feminist allies to be able to get into women's beds in a sneaky manner.
Tolerance of other views (hearing them without screaming or trying to get the speaker fired, etc.) correlates strongly with the level of confidence the listener has in his own views, I think. Anyone who is secure in his own beliefs and actions has no need to fear contrary perspectives, and is content to answer them with reason, not force.
I think you're right to an extent, but I also think it has to do with curiosity and open-mindedness. As a woman and someone who is not entirely confident in my views about any one issue (I am confident in my principles, though), I would never shout down or try to cancel a speaker because I know that by allowing them, as well as their opponents, to speak, my views will be informed and perhaps more solidified as a result.
If you don’t trust in the reason of your fellow citizens you don’t trust them coming up with the right answers
Another essential quality is intellectual humility, recognizing that you cannot expect to understand everything, or even just another human being, so you know that there is value in listening to people with a different perspective.
Hi Chapin! Fascinating stuff but could you point me to where the raw data are? I couldn't find them at the link you gave, and the heat maps are frankly giving me a headache :)
The raw data is here: https://rankings.thefire.org/methodology
The tolerance values are calculated as even sums of the left/right speaker questions and scaled 0-100%. Let me know if you have other questions about the data/my analysis.
I can't find these charts or their data in the reports. Could you please pinpoint them? I want to know what is meant by "tolerance" in this context.
Tolerance is measured by asking whether students would allow hypothetical controversial speakers to speak on campus. It's explained in the longer article, and the statements are listed in the addendum:
https://eternallyradicalidea.com/p/men-are-more-tolerant-of-the-other
https://docs.google.com/document/d/1QG2qjqd-gKj9SFIkAeS4wtZ2q_jQwUVCjz9d-k-dPZk/edit?tab=t.0
Most of the American cancel culture participants are nothing more than the modern equivalent of spinster church ladies. They love to point out the flaws of others and point out their own virtue. In the absence of public morality they now use more fashionable metrics such as bigotry. This is most obvious in the "me-too" accusations where they struggle so hard because consent is the only concept left to exploit.
Now that people stopped attending church, the busy bodies moved on to becoming university administrators and high school teachers. They LOVE judging others. They ruin every industry they come to dominate, then blame everyone else for running away. Even the women who hate this have trouble recognizing it for what it is. instead they say "I hate working for women" when in reality they hate being around close-minded hypocrites who happen to be female.
The only way to fix this is to legislate it, and to promote rules-based environments while dismissing and demonizing taboo-based environments. The reason so many places are so oppressive is because these busy-bodies love to enforce taboos, and they hate rules because they are explicit and cannot be warped in order to attacks others.
Tolerance is essential in a multicultural society. We are careening towards civil war. We must fix broken environments before the rot spreads to important institutions.
Do I understand properly that libertarian women are classified as left-biased and men as right-biased?
Thankfully we have organizations like Braver Angels and BridgeUSA that are working to cultivate tolerance and respectful disagreement. I just wrote Substack articles about both of their efforts; I hope they become more mainstream.
https://spiritandsword.substack.com/p/is-the-bridging-movement-the-counterculture
The fact that only one of the issues directly affected women doesn't exactly jam up that argument. It would also make sense if, being historically more constrained by the consequences of political decisions, women are generally more reactive on the behalf of whatever group is most affected while men get to remain aloof and treat political discourse as a kind of game. I suspect this is at least part of a larger explanation.
I think women tend to value feelings/emotions more. They don't wanna see people get their feelings hurt. And some opinions hurt people's feelings, sooo...
Of course, by censoring speech, they are creating anger, frustration, and resentment.
Men will just get violent to shut them down.
We have rules in place to deal with male rage and violence. We do not have any structures or policies in place to deal with female shaming and reputation destruction. Besides, male violence peaks in their teens and is no longer needed once they are out in a world where one's status is defined.
Men are no more violent today than they were in the past. Our rules can handles that, at least when we want them to. Of course, now many cities have decided to stop punishing violent men. Want to guess which sex was pushing for that change?
"Men are no more violent today than they were in the past"
And if anything men in Europe and east Asia are far less violent than in the past because we selectively executed about 1% of the male population each generation (largely psychopaths) for a very very long time (tribal societies did not do this, they instead rewarded the most violent men with harems).
Yeah, you guys are doing a great job
I do not understand your comment. Are you saying we should stop enforcing crime? Should we legalize rape and domestic violence? I cannot understand where your comment came from, unless it is a bot.
Male reading comprehension skills ftw
Your entire schtick is to try to provoke with male bashing. It deserves a Mencken horse laugh.
Guess who's more likely to want harsher punishments for violence. Men, or women?
You are the one not showing comprehension.
Banned.
If the difference is simply that women prefer more social harmony, wouldn’t we expect to see this effect across generations? Has anyone done that study? I’d bet Gen X and Boomers have less of this effect and if we could find enough of the Greatest Generation still, we wouldn’t see this gap at all, or it might not even reach the level of significance. There are other things going on with millennial and Gen Z women: they have been educated by teachers and professors to despise debate and have been overwhelmingly captured by those views, while those views have actually driven men away, in large part because the package of views includes the concept of “toxic masculinity”. Another potential contributor to the effect could be that young women are far more engaged on social media than young men are. Young men are more engaged with online video games. Social media literally trains your brain to react in strongly antisocial ways to things you disagree with, due to the internal reward system of clicks, likes, and shares.
We see the gap in all generations. Female professors and students are both far more likely to support censorship in academia than their male equivalents.
Women were also the staunchest supporters of totalitarian movements in other countries.
But both the professors and the students only make it into their jobs and placements if they parrot a suite of leftist views which includes support for censorship. So a sample of only people involved in academia is extremely skewed. In order to determine whether women are truly less tolerant in significant percentages, a good study would include a broader sample of men and women from across society, not just from academia.
Our national data is less comprehensive, but shows a large gender gap too:
https://expression.fire.org/p/conservative-women-are-as-likely
Thanks for this! Very interesting 🤔 Do you know if anyone has looked at whether social media use could be covariant with views on speech?
Women did not dominate certain fields in the past. We are in a brave new world now where industries such as book publishing, journalism, university administration and others are now dominated by women for the first time in human history. It is in these fields where we see the decline in tolerance.
Interesting evidence to a fairly common anecdotal claim.
I've theorized for some time now that this is potentially linked to direct experience of the consequences of physical violence in interpersonal disputes.
Little boys grow up with much more experience in this area than little girls. Hence, they think differently about the consequences of escalating conflict. Women, being more inclined to covert or indirect forms of violence, rarely feel the sting of retributive conflict escalation. Engaged by proxy, it will be men who bare the burden of this socially.
Your plotted outcomes appear to belie the idea that there is no difference between left and right on this - after controlling for sex, the right-wing show more tolerance than the left (according to the plot, but not sure if this is statistically significant). This effect is smaller than the sex effect, but it looks like there is some effect. You should investigate and revise your article accordingly.
Dear Chapin,
Thank you for the interesting article.
I took some time to look into the analysis underlying your post because I wanted to understand precisely how the claim that “men are more than three and a half times more likely to be perfectly tolerant than women” was derived.
To do so, I read the subsequent discussion in the comments in which you indicated the data source and the methodology, and downloaded and re-analyzed the publicly available raw data from the 2026 College Free Speech Rankings that you used.
One suggestion up front: I think the methodology should be set out explicitly in the article itself rather than being left to the comments section. I spent a considerable amount of time trying to reproduce the result, and without the methodological clarification it is very difficult to understand how the headline claim is obtained from the underlying data.
My understanding is that the analysis is based on responses of men and women to the following survey item:
"Student groups often invite speakers to campus to express their views on a range of topics. Regardless of your own views on the topic, should your school ALLOW or NOT ALLOW a speaker on campus who has previously expressed the following idea?"
The six statements are:
"Transgender people have a mental disorder."
"Abortion should be completely illegal."
"Black Lives Matter is a hate group."
"The Catholic Church is a pedophilic institution."
"The police are just as racist as the Ku Klux Klan."
"Children should be able to transition without parental consent."
I downloaded the raw male and female datasets, applied the survey weights (School Universe Weight), excluded missing values, and calculated support as: "Probably should allow" + "Definitely should allow"
Using that methodology, I obtained the following results:
Statement Women allowing (%) Men allowing (%) Difference (percentage points)
"Transgender people have a mental disorder." 17.5 39.1 +21.6
"Abortion should be completely illegal." 30.9 53.9 +23.0
"Black Lives Matter is a hate group." 15.6 36.7 +21.1
"The Catholic Church is a pedophilic institution." 31.0 42.3 +11.3
"The police are just as racist as the Ku Klux Klan." 30.3 40.8 +10.5
"Children should be able to transition without parental consent." 46.3 48.7 +2.4
These results suggest that men are consistently more likely than women to allow the speakers in question, but that the size of the gap varies substantially by topic, from 2.4 to percentage points to 23 in the case of someone having a track record of supporting the full ban of abortion.
What puzzled me is that I could not reproduce the "3.5 times" finding using what I understand to be the actual tolerance scale. If tolerance is calculated, as described, by combining the speaker questions into an equally weighted index and scaling it from 0 to 100, I obtain approximately:
Women: 33.6
Men: 45.9
In other words, men score about 12 points higher than women on the tolerance scale, or roughly 1.37 times higher, not 3.5 times higher.
The only way I was able to reproduce a figure close to the one reported in the article was by constructing a separate category of "perfect tolerance", defined as respondents who answered "Definitely should allow" on all six questions simultaneously. Under that definition, I obtain approximately:
Women: 1.9%
Men: 6.8%
which yields a ratio of about 3.6:1.
If that is indeed the methodology used, I think it is essential that it be stated prominently in the article. The phrase "perfectly tolerant" strongly implies that the result is a direct measure of tolerance, whereas in reality it depends entirely on a very specific coding decision requiring the maximum possible response on six separate questions.
My concern is that the article currently invites readers to conclude that the data show men to be generally 3.5 times more tolerant than women. However, the underlying data seem to support a much narrower and more nuanced conclusion: men are more likely than women to allow the specific speakers included in the survey, with the magnitude of the difference varying considerably across issues.
Put differently, the headline appears to rest not on the tolerance scale itself, but on a highly restrictive binary classification applied afterward. Because only 6.8% of men and 1.9% of women satisfy that criterion, the resulting ratio is largely a product of that methodological choice rather than a straightforward summary of the underlying data.
Could you thus kindly clarify whether the 3.5x figure was derived from the requirement that respondents answer "Definitely should allow" on all six items, or how else you came to this finding.
Given how central these choices are to the article's conclusion, I think greater methodological transparency would substantially strengthen the analysis and help readers evaluate what is actually being measured.
Thank you very much and have a nice day!
Thanks for your comment. Yeah, that's exactly how I got the 3.5 figure.
In my articles, I'm always balancing trying to be as detailed and transparent about my methodology as possible and trying to be accessible to people without technical backgrounds. Not as easy task, and I'm always open to suggestions on how to do so better. I would've guessed that saying "men are more than three and a half times more likely to be perfectly tolerant than women" would be fairly clear. (I can't think of anything way to interpret "perfectly tolerant" besides "would definitely allow all speakers". Can you?) Though I've seen others misinterpret that line as meaning "3.5x more tolerant".
Also, the 3.5 figure, while notable, isn't central to the article's main conclusions. The central result of the article is that men often have grater tolerant towards speakers on the other side than women do towards speakers on their own side, which comes from the findings that slightly conservative men being more tolerant of left-wing speakers than democratic socialist women, a slightly liberal men being more tolerant of right-wing speakers than somewhat conservative women, etc.
See the longer version of the article, which notably has heatmaps where perfect tolerance is the top right square:
https://eternallyradicalidea.com/p/men-are-more-tolerant-of-the-other
Also see the addendum to that article:
https://docs.google.com/document/d/1QG2qjqd-gKj9SFIkAeS4wtZ2q_jQwUVCjz9d-k-dPZk/edit?tab=t.0
And my code and data are here:
https://github.com/chapin-lenthall-cleary/firearticles/tree/main/gender_tolerance
I am going to say it. Nowhere is this more glaring than with the issue of abortion. When the word is spoken, the sword drops on the table dividing women, erasing men and galvanizing rhetoric. Unfortunately for many women, their feet are in concrete on this issue. It becomes the base of their political views that they can not be moved from it.
Exactly this is what makes them poor politicians, and citizens that base their vote on one issue.
This essay presents itself as empirical clarity, but it repeatedly turns a narrow survey result into a moral judgment the data does not support.
What the data actually shows is straightforward: male and female students answer a specific survey question differently. Men are more likely to say they would allow controversial speakers on campus; women are more likely to say they would not. This pattern holds across political ideology, and the gender gap is larger than the ideological gap. That is the finding.
Everything beyond that is interpretation.
The survey measures tolerance as willingness to allow hypothetical speakers in an abstract, consequence-free setting. It does not measure openness to debate, reasoning ability, emotional stability, or authoritarian impulse. It does not measure how people behave when speech has real institutional, social, or personal costs attached. Treating this metric as a proxy for character is unjustified.
The essay nonetheless frames male responses as civic virtue and female responses as censorial instinct. That move is emotional, not empirical.
From the beginning, tolerance is treated as an unqualified good and refusal as a failure. But the data never weighs error costs. It never asks whether some refusals are rational, preventative, or responsive to real-world asymmetries. Those assumptions are simply smuggled in.
The dismissal of alternative explanations is similarly weak. The author argues that women aren’t responding differently because they’re more affected by speech, since only abortion is explicitly sex-linked and the gap doesn’t spike there. This misunderstands how risk works. Speech-related costs for women aren’t limited to specific policy topics; they include higher exposure to harassment, reputational penalties, social sanction, and uneven institutional protection. These are ambient conditions, not speaker-specific variables, and the survey cannot detect them.
There’s also a clear narrative asymmetry. Men’s responses are treated as the neutral baseline. Women’s responses are treated as a deviation that requires explanation. That is not a methodological necessity; it’s a framing choice.
The most revealing moment is the speculative turn: the suggestion that women “prefer social harmony.” Presented as modest conjecture, it functions as character inference. Harmony becomes conflict aversion, which becomes censorship, which becomes a threat to democratic discourse. None of those steps are supported by the data. They are layered assumptions.
What’s entirely missing is any discussion of the costs of excessive tolerance. The essay treats tolerance as costless and censorship as uniquely dangerous. But tolerating speech and governing institutions are not the same task. Universities are not abstract marketplaces of ideas; they are organizations with duty-of-care obligations and power asymmetries. A respondent who weighs those realities more heavily will score as “less tolerant” by design.
By the end, the essay stops analyzing data and starts issuing moral warnings about emotional comfort undermining society. At that point, the data is no longer being interpreted; it’s being recruited to serve a worldview.
A more accurate summary would be simple: men and women answer abstract questions about allowing controversial speakers differently, and this difference outweighs ideological variation. Explaining that gap requires assumptions about harm, risk, and institutional responsibility that the data itself cannot resolve.
The data is real. The emotional story built on top of it is not.
lol. Awful ai-slop.